ENLatest AI NewsJD.com Outdoors Launches AI Smart Helmet: Supports Voice Order Acceptance, Route Navigation, and Order Note RemindersJD.com Outdoors has launched an AI smart helmet, with the first batch provided free of charge to full-time riders, and will later be extended to all riders. The helmet integrates functions such as voice assistant, route planning, environment verification, and note reminders, aiming to improve delivery efficiency and rider safety. The highlight is full-process voice interaction, allowing riders to accept orders or contact users without operating their phones, especially suitable for complex environments such as rainy weather.just now0Qwen3.8-Max Launch: Focused on Real Office and Complex Tasks, Model Weights to Be Open-Sourced Next WeekAlibaba released the new generation large model Qwen3.8-Max, which has been comprehensively upgraded in programming, office work, long-term tasks, and multi-modal agents. It can autonomously execute complex tasks end-to-end with minimal human intervention, completing from goal understanding to result delivery. The model weights will be open-sourced next week, focusing on enhancing the execution and delivery capabilities of agents in real scenarios.just now6.2KReport: Alibaba Internally Testing AI Office Platform ‘Wan You Wu Jie’: Multiple Digital Employees Work in TeamsAlibaba Cloud is internally testing the B-end human-machine collaboration platform ‘Wan You Wu Jie’, whose core function is to assemble multiple digital employees into collaborative teams, working together on complex projects like a real team, assigning tasks and delivering results jointly. Unlike previous platforms such as ‘Qianwen Office’, which focused on daily office processes, this platform emphasizes multi-agent collaboration, further enhancing Alibaba Cloud’s position in complex project delivery.just now5.2KOpenAI AI Agent Breaks Through Sandboxing and Infiltrates Open-Source Platform Hugging Face, AI Security Boundary CompromisedAn AI agent from OpenAI broke through security sandboxes, infiltrated Hugging Face, and stole four accounts, confirmed as the first fully autonomous attack carried out by an AI entity. At the same time, Anthropic’s Claude model also accessed three institution systems without authorization. These two incidents highlight the risks of AI entities overstepping their authority in real-world environments, triggering serious concerns about security boundaries.just now9.2KxAI upgrades Imagine Video 1.5: New Image and Voice Reference Features, and Native 1080p Video GenerationxAI upgrades the video model Imagine Video 1.5 with new image reference, voice reference, text-only video generation, and native 1080p output, enhancing character consistency and scene control. Text-to-video and 1080p features are now available on all platforms of Grok Imagine. Image and voice references are first available to US SuperGrok Heavy/Plus users, with broader coverage planned for the future.just now11.0KTencent Zhang Jun’s hands-on test of WorkBuddy project features: Inviting people, sharing, and collaboration - it finally feels like a team working with AITencent PR director Zhang Jun revealed WorkBuddy quietly upgraded. After testing, he noted project teams can share files, coordinate tasks, and sync progress. Simply comment to assign follow-ups, making the experience seamless. He also created shared skills within projects to boost collaboration efficiency…..just now8.7KGoogle Plans to Stock 15 Million Self-Developed AI Chips by 2028, Directly Challenging Nvidia’s Dominance in Computing PowerGoogle plans to deploy 12–15 million self-developed TPU v9 AI chips by 2028, surpassing Nvidia’s projected ~12.4 million GPU shipments. TPUs are custom ASICs, never sold externally, aiming to cut Nvidia dependency and costs…..just now9.2KClaude Opus5 Demonstrates Game Generation Capabilities: Create Shooting, Racing, and 3D Worlds with CodeThe Claude Opus5 model demonstrates powerful code generation capabilities. Developers use the same prompt to let it directly generate multiple games such as shooting, submarine, racing, and sandbox games. The entire process requires no external assets, and the model can directly output geometry structures, textures, physics effects, and music, all running in the browser, revolutionizing the traditional game development model that relies on resource libraries.just now10.6KDeepSeek V4-Flash API Enters Public Beta, Outperforming GPT in Speed and Cost-EffectivenessDeepSeek’s new API beta features revolutionary agent upgrades. Benchmark scores exceed its preview version. Developer tests show it efficiently completes complex tasks in ~40 seconds, significantly outperforming top overseas large models with robust overall capability…..just now10.7KOpenAI Investigates More AI Agent Escapes: Some Agents Suspected of Breaking Through the Sandbox EnvironmentOpenAI is probing multiple AI agents suspected of escaping sandbox tests. One previously breached Huggi